PROJECTION OF FLOOD SEASONALITY CHANGES IN A GARHWAL HIMALAYAS RIVER BASIN DUE TO GLOBAL WARMING

Abstract

The impact of a warming climate on snow- and rain-dominated river basins constitutes a significant research challenge and the potential for severe socio-economic risk. The combination of the snowmelt and rainfall-induced flood peaks in the Garhwal region of India is of hydrometeorological and hydrographic interest. This study focuses on the projection of seasonality changes in floods in a Garhwal Himalayas basin under global warming. The research in this context is somewhat uncertain in the proposed study area of the Himalayas, mainly due to the scarcity and unavailability of long-term and high-resolution meteorological data in that region. But after setting up Automatic Weather Stations and Gauge and Discharge sites in the Garhwal region in 2016, the observed data of the past five years lay the basis for understanding the different flood generating regimes. In the present study, the hydrometeorological data of two sub-catchments in the Bhilangana River basin, i.e., the sub-catchment at Sarasgaon and the sub-catchment at Ghansali, have been analyzed in terms of available observed data and projected future climate data of rainfall and temperature. We have analyzed the IMD historical maximum monthly rainfall (1901-2020) over the study region and found evidence of shifting of maximum rainfall peak backward up to June, thus the possibility of the seasonal changes in the flood peaks under warmer climate scenarios (if experienced in future). We also compared the different precipitation datasets available with respect to the observed data at daily, monthly, quarterly, and yearly time scales. Those data are crucial for any analysis of possible changes in seasonal hydrometeorological conditions. We found that the IMD precipitation dataset best matches the observations. The projected climate ensemble of the chosen dataset (NEX-GDDP) required significant correction with respect to observed data to counter underestimation. Therefore, we have used quantile-based mapping to adjust the biased projected climate dataset of NEX-GDDP. Also, the corrected projected precipitation and temperature of time window 2071-2099 of RCP 4.5 and 8.5 scenarios are significantly higher than that of the corrected Historical time window 1971-1999. To project the changes in floods for past and future scenarios, we attempted hydrological modelling of the catchment at Sarasgaon (with five years of observed data available) via the CemaNeige_GR4J model in the airGR package in R software. Before this, we also gap-filled the discharge time series of the Sarasgaon gauge station by developing the rating curve.